Distributions, Volumes, and Compositions

Distribution of Composite Cones in Volcanic Arcs

Volcanic arcs may break into segments, each of which is parallel to the arc trend and contains between two and a dozen volcanic centers (Stoiber and Carr, 1973; Marsh, 1979a; Carr et al ., 1982). Arc segments have been identified by mapping offsets in lines of volcanoes, contrasting volcano shapes and eruption styles, clusters of small basaltic volcanoes located behind the volcanic front near breaks, transverse fault zones, and clusters of large, shallow earthquakes at the segment boundaries (Carr et al ., 1982). Segmentation has been well-documented in Central America (Carr et al ., 1982), the Aleutians

Fig. 7.2
Simplified two-dimensional diagram of a magma supply  vs  percolation rate model of lithospheric
magmatism. Axes are schematic and depict only the relative magnitudes of supply and the
modification of primitive magma in the systems shown. Asterisks indicate systems thought to be
characterized by the rapid transient injection of basalt at restricted crustal levels, which results in the
generation of rhyolite but little intermediate magma. In contrast, most continental and orogenic systems
may involve a diffuse injection throughout much of the lithosphere as well as subsequent mixing and
crustal mobilization, which in turn produce chiefly magmas of intermediate composition.
The degree of potential magma modification increases with increasing crustal thickness,
compositional contrast, and magma residence time.
(Adapted from Hildreth, 1981.)

(Marsh, 1979b), and the Mariana-Volcano Islands (Meijer, 1982).

Figure 7.4 shows a volcanic arc in Central America where zones interpreted as segments range in length from 55 to 260 km. Table 7.1 presents a comparison of segment length and volume for Central American volcanic arc segments. Stoiber and Carr (1973) reported that the volume of erupted lavas and pyroclastic rocks per kilometer increases with increasing segment length, from 1 km³ /km for the shortest (55 km) to 5.2 km³ /km for the longest (260 km). It has been proposed that arc segments reflect breaks and uneven surfaces within the descending plates (Carr et al ., 1982; Marsh, 1979a, b). Segment boundaries commonly

Fig. 7.3
Schematic cross-sections that depict two possible stages in the development of igneous systems in which tectonic
extension is subordinate and shallow. This model applies to (a) early and (b) intermediate stages of island arcs,
continental margin arcs, and mid-continental igneous systems.
Nearly all the heat for this system is supplied by basalt injection.
(Adapted from Hildreth, 1981.)

coincide with grabens that are oriented perpendicular to the plate boundary. Burkhart and Self (1985) interpreted these grabens as expressions of tectonic extension—rather than segmentation—of the plate.

The distance between composite cones located within adjacent arc segments is fairly uniform, and the volume of material erupted from each center is approximately proportional to that distance (Marsh, 1979a, b). For the Aleutian Island arc, this spacing is ~70 km (Marsh and Carmichael, 1974). Meijer (1982) noted that along the Mariana volcanic chain of the western Pacific, the spacing between volcanic centers ranges from 20 to ~80 km; this spacing is correlative with volcano size: the small volcanoes are the most closely spaced. Within segmented arcs, small, usually monogenetic cones may form at a distance of ~50 km behind and parallel to the volcanic front. These less voluminous cone clusters may develop 3 to 4 m.y. after the beginnings of arc volcanism (Marsh, 1979a).

Volcanic Eruption Rates and Relative Volumes for Magma Types in Composite Cones of Volcanic Arcs

Volcanic arcs and their composite cones contain a full spectrum of magma compositions—from basalt to rhyolite. Temporal variations in these compositions provide clues about the depth and size of the intrusive rocks that were their thermal sources. Compositional variations are affected by the rate of plate movement, the angle of plate descent, irregularities in the descending plate, crustal thickness, and the depth and residence time of the magma reservoirs.

Fig. 7.4
The segmented volcanic front of Central America, in which active volcanic fields are shown as
shaded areas. Stippled vertical bars mark the transverse breaks in the arc. The thin, parallel lines
mark depths to the inclined seismic zones (contour interval is 50 km); offshore 1000-m
contours are depths below sea level.
(Adapted from Carr et al ., 1982.)

Another factor affecting the longevity of thermal sources is the extrusion rate for individual volcanoes and volcanic chains. This factor is difficult to evaluate because of the buried eruption sequences, erosion, and widespread distribution of pyroclastic products, but a number of studies have provided enough data to allow general estimates. Nakamura (1974) and Crisp (1984) reviewed data for volcanic output on a global scale and found that subduction-zone-related volcanoes produce from 0.4 to 0.75 × 10⁶ km³ /m.y. Working with individual arcs, McBirney et al . (1974) and Sugimura and Uyeda (1973) concluded that the volume of material erupted for the Cascade Range and Japan, respectively, was ~5 km³ /m.y./km of arc. These general estimates were confirmed in a more detailed work by Sherrod and Smith (1990), who found that the extrusion rates in arc segments of the Quaternary Cascades volcanic arc range from 0.21 to 6 km³ /m.y./km of arc. Variations in the volume of material erupted from volcanoes of the Lesser Antilles and Central America may be related to both crustal thickness and rates of plate convergence; over the last 100,000 years, production rates have been 3.1 km³ /m.y./km of arc in Central America and 4 km³ /m.y./km of arc in the Lesser Antilles (Wadge, 1984). This relationship between convergence rate, crustal thickness, and magma types is shown in Table 7.2.

If an intermediate or silicic magma body—either small or large—is to rise buoyantly to crustal depths, it must be heated from below by basaltic magmas from the asthenosphere. Without this thermal boost, silicic magma chambers cool and solidify; they may never reach the upper crust (Lachenbruch et al ., 1976; Eichelberger, 1978). Fractionation and mixing of basaltic and silicic melts can produce the spectrum of magma types seen in composite cones. These compositional variations are controlled by the rate of magma supply, crustal thickness, rate of magma percolation through the crust, and extrusion to intrusion ratio (Figs. 7.2 and 7.3; Hildreth, 1981).

To calculate the number of shallow crustal magma bodies that might provide heat to geothermal systems, it is necessary to determine the relative volumes of magma types and their ages for each composite volcano and, if possible, for an entire arc. Central Cascade Range volcanism in North America began in the earliest Pleistocene with the eruption of widespread basaltic cones and flows and the construction of overlapping shield volcanoes (McBirney and White, 1982). Activity became more localized at centers from which more andesitic lavas and tephra were erupted. This activity formed the base upon which the large composite cones were constructed during the past million years. The volume measurements of

McBirney et al . (1974) indicated that most of the province consists of basaltic scoria cones and lava flows and that andesitic composite cones make up only 15% of the erupted material. Studies of Mount Jefferson indicate that the early basaltic activity produced >100 km³ , but that the cone-building stages involved only 25 km³ of andesitic magma (Fig. 7.5).

Clark (1983) found that the Three Sisters volcanic complex of the southern Cascade Range was erupted onto a broad base of basaltic lavas (57 km³ ), which included a much larger relative volume of andesite (30 km³ ) and rhyodacite-rhyolite (3.5 km³ ). Rhyodacitic domes erupted during the past 2300 years from aligned vents that cross the summit regions of the complex. These eruptions may have tapped only a small volume of a compositionally zoned, shallow magma chamber of much greater volume (Scott (1987).

By plotting volumes of erupted material vs SiO₂ compositions for the Quaternary volcanic rocks of Japan, Aramaki and Ui (1982) demonstrated that changes along the arc may be related to both plate movement and crustal thickness. Arc segments that consist of composite cones, domes, and calderas are mostly andesite-dacite-rhyolite—all of which are indicative of shallow crustal magma bodies. Most of the rhyolitic materials are associated with large calderas. The basaltic segments are made up of pre-dominantly simple cones and lava flows (Fig. 7.6).

Fig. 7.5
Relative volumes of rock types associated with
Quaternary composite cones in the Cascade
Range of North America. (a) Mount Jefferson was
constructed during four main periods of activity;
rock volumes vs SiO₂  content for each stage are
shown in the graph. (b) Three Sisters complex,
which is characterized by more siliceous rocks
than Mount Jefferson. Of the two, the Three
Sisters complex has more potential as a
geothermal resource.
(Adapted from McBirney and White, 1982.)

Inferred Intrusive Volumes and Their Depths below Composite Cones

Using a model originally developed for evaluating large-volume eruptions of silicic magmas and subsequent caldera collapse, Smith and Shaw (1975) determined that the volume ratio of magma chambers to erupted material is ~10:1. This conclusion was based on models of magma transport into the Earth's crust, exhumed intrusive-volcanic complexes, petrologic indicators, and geo-physical studies of active igneous systems. The authors later applied this model to the evaluation of all the high-grade geothermal systems in the U.S. Shaw (1985) took this model even further when he calculated volume-periodicity relations for explosive eruption activity in a variety of volcanoes—from composite cones located along plate boundaries to large mid-continental calderas. Most of the magma chambers below andesitic-dacitic cones examined in this study are located within the upper 4 km, and several are within 2 km of the crust.

Crisp (1984) and Wadge (1984) approximated intrusive-to-extrusive ratios for all volcano types, including composite cones and the associated domes and calderas along subduction zones (Table 7.3).